Sharing magnetic resonance imaging (MRI) brain scan results among colleagues is often difficult, because the data sets involve large numbers of digital images and require specialized software for effective analysis and visualization. Detailed analysis of anatomical MRI scans, for example, (each containing hundreds of images) benefits by using sophisticated software to coordinate and display multiple views of the same tissue, to segment out tissue structures of interest, to reconstruct 3-D surface views from serial tissue slices, or to re-slice image data sets along different planes. Analysis of functional MRI scans, (each containing tens of thousands of images) requires additional specialized statistical image processing software to analyze and visualize task-specific spatial and temporal fluctuations in MR image intensity. Providing shared access to such data sets and analysis tools is a problem for both clinical and research brain scans. Another problem in MRI analysis is that simply seeing brain images is not enough; in most cases it is also necessary to identify particular brain areas by name or reference coordinates. Given the complexity and variability of brain anatomy, as well as variability in image slice position and orientation, identifying individual areas can be quite difficult. Specialized software, this time in the form of an interactive reference brain atlas, could greatly facilitate the tissue identification process. Once identified, linking that identification to the user's MRI images would significantly enrich the data set for viewing by others. Here we propose to develop a web-based image visualization and analysis system for brain MRI images that should greatly improve the portability and accessibility of large data sets, without compromising confidentiality or statistical analysis capabilities. Our solution brings together expertise in MRI image processing and web-based brain atlas and neuroanatomy software to create ICARUS, the Interactive Computation and Atlas Remote-User System. In this Phase I SI7R project we will produce and demonstrate a prototype of this software package.